Add more Edge->TOSA lowerings and fixes (#475)

Summary:
* FP32 Softmax lowering
* Conv2d lowering with groups
* Add promtote_shape and tranpose helpers for aligning with TOSA rank requirements
* Add more e2e test cases
* Other minor fixes

Pull Request resolved: https://github.com/pytorch/executorch/pull/475

Reviewed By: cccclai

Differential Revision: D49624396

Pulled By: digantdesai

fbshipit-source-id: b0db46d087f831b4f348d73306b77c934127707f